1. Field of the Invention
This invention relates to electrical connection systems, and, more particularly, to card edge connection systems wherein conductive pads arranged in rows adjacent to an edge of a daughter card are engaged by rows of contacts within a connector mounted on a mother board.
2. Description of the Background Art
In the construction of computers and other electronic devices, a need for modularity and design flexibility has made it necessary to build many devices using combinations of various circuits formed on individual printed circuit cards. Electrical connectors provide the means required for the removable assembly of such cards, including the circuit interconnections required among them. Industry trends, such as the miniaturization of electronic components and concurrent reductions in the cost of providing many functions through the use of electronic circuits, are greatly increasing the density of circuit lines on many printed circuit cards, placing similar increased density requirements on electronic connectors.
Many computers today use what is called a "mother board, a "system board," or a "backplane" board, which includes a number of circuits leading to electrical connectors, into which a number of "daughter boards" or "adapter cards" may be plugged to personalize a specific system and to provide particularly for connection to peripheral devices. In some computers, processor circuits on circuit cards are plugged into connectors in this way, as well.
For the sake of simplicity, in the following discussion the assumption is made that a "daughter card" is plugged into a connector attached to a "mother board" by soldering, that the mother board lies horizontally, and that the daughter card extends upward from the connector. It is understood that changes in orientation can easily be made without varying the concepts or hardware involve.
One type of connector which has proven especially effective in the construction of electronic devices is the card edge connector, which is typically configured to removably receive an edge of a printed circuit card in a central card-receiving slot. The card includes a single row of conductive pads on each side adjacent to the edge inserted in the connector. The connector includes a single row of flexible spring contacts on each side of the central slot, configured to be deflected outwardly by the insertion of the card and thereafter to make contact with the pads on the card. These contacts extend as solder tails outside a surface of the housing opposite to the central slot, to be fastened by soldering to individual circuits in the mother board. Thus, electrical connections are formed between circuits attached to the conductive pads on the card and circuits within the mother board. Examples of card edge connectors of this type are found in U.S. Pat. No. 3,868,166, which was issued to J. P. Ammon on Feb. 25, 1975, in U.S. Pat. No. 4,795,374, which was issued to P. L. Richworth et al. on Jan. 3, 1989 , in U.S. Pat. No. 4,846,734, issued to Lytle on Jul. 11, 1989, and in U.S. Pat. No. 4,891,023, which was issued to J. E. Lopata on Jan. 2, 1990.
Furthermore, the use of multiple rows of contacts on each side of a card edge connector, providing contact surfaces along the central card-receiving slot at various distances from the entrance of this slot, together with a corresponding pattern of parallel rows of contact pads on each side of a daughter card, is practical and has been described, for example, in U.S. Pat. No. 4,806,103, which was issued to W. Kniese et al. on Feb. 21, 1989, and in U.S. Pat. No. 4,934,961, which was issued to H. Piorunneck et al. on Jun. 19, 1990.
Where there are different types of hardware that can be interconnected, with the possibility of causing malfunctions or damage if the wrong combinations are chosen, combinations of keys and keyways are often used to prevent misconnections. Typically these combinations are used simply to prevent the connection of certain devices and to assure that connected devices are properly oriented with respect to each other, as described, for example, in U.S. Pat. No. 4,257,665, which was issued to H. John et al. on Mar. 24, 1981, in U.S. Pat. No. 4,376,565, which was issued to P. S. Bird et al. on Mar. 15, 1983, in U.S. Pat. No. 4,715,820, which was issued to H. W. Andrews, Jr. et al. on Dec. 29, 1987, and in U.S. Pat. No. 4,884,975, which was issued to L. Peizl, et al. on Dec. 5, 1989. Key and keyway arrangements can also be used to determine how far a connector is inserted into a board in which it will be soldered, as described in U.S. Pat. No. 4,479,686, which was issued to M. Hoshino et al. on Oct. 30, 1984, or to determine how far a daughter card is inserted into a connector, as described in U.S. Pat. No. 4,934,961 to Piorunneck.
One particular concern with various types of changes to hardware is that of upgradability. In the marketplace, this concern is often expressed as a need to protect the investment of the customer. It is desirable that, when a customer purchases a new system, he should be able to use as much of his old hardware, which can represent a significant investment, as possible. On the other hand, if the desire for this kind of compatibility is allowed to dictate the way a new system is designed, progress often cannot be made toward increasing system function and performance. In the area of electrical connectors, increased function often means that more signal lines will be required, and that connector line densities must be increased. Therefore, it is particularly desirable to provide a means for increasing the number of signal lines through a connector in such a way that both circuit cards having an improvement using more such lines and older circuit cards with fewer lines can be installed.
This type of compatibility is achieved by using the type of connector described in U.S. Pat. No. 4,934,961, to Piorunneck et al., which describes a bi-level card edge connector having two rows of contacts on each side of a central card-receiving slot. A new type of card configured for use with this connector has two rows of conductive pads on each side--a lower row adjacent to the card edge and an upper row adjacent to the lower row. The connector includes several keys extending across the central slot. When this new type of card is inserted into the slot, the keys align with keyway slots in the card, allowing the card to be fully inserted, so that, on each side, an upper row of connector contacts makes electrical connections with an upper row of conductive pads on the card, while a lower row of such contacts makes electrical connections with a lower row of conductive pads on the card. An older type of card, which has only a single row of conductive pads, adjacent to the edge, also lacks these keyway slots, so that such an older card can only be partially inserted, leaving the upper row of connector contacts providing electrical connections with the only row of card conductive pads on each side, while the lower row of connector contacts makes no electrical connections. Thus, the contact density of a card-edge connector is increased while means are provided to allow the use of both cards of the new type, having two rows of conductive pads per side, and of the old type, having only one row of conductive pads per side.
However, the connector and daughter card described in this patent, U.S. Pat. No. 4,934,961 to Piorunneck et al., does not have the advantage of compatibility in the other direction. Since an interconnection has been established wherein certain daughter cards have about twice as many connected circuits as others, and wherein both types of daughter cards can be used with the connector, the additional circuits must be non-essential, while essential circuits are connected through the only row of contact pads in a daughter card of the old type. Since these circuits are connected through the upper rows of connector contacts, these contacts must be the ones connected to essential circuits within the mother board, while non-essential circuits are wired through the lower rows of connector circuits and contact pads. If a card of the new type, having two rows of contact pads on each side, is plugged into a connector of the old type, only the non-essential circuits of the lower rows of pads will be connected with the essential circuits of the only row of contact springs. Thus, while the interconnection system proposed in this patent meets a need for being able to plug a card of an old type into a connector of a new type, it does not meet a need for being able to plug a card of a new type into a connector of an old type.
A number of connector designs are used to provide a "zero-insertion-force" feature, which allows the insertion of a daughter card into a connector without dragging the contact pads of the across the connector spring contacts. This feature can be achieved by stamping or forming the spring contacts so that they do not contact the daughter card in their free, undeflected state, i.e. by making the opening between opposing contacts wider than the width of the card. An actuator is provided to push inward on the spring contacts after the insertion of the card so that they are forced against the contact pads. Typically, a separate movement, to move the actuator, is required of the person assembling the daughter card into the connector.
An example of this kind of connector design is found in U.S. Pat. No. 4,080,027, which was issued to J. E. Benasutti on Mar. 21, 1988, which describes a connector wherein each contact terminal has a rigid conductive body with an attached flexible section in turn made up of an insulative portion and a conductive portion. After the insertion of a daughter card, a flexible actuator extending between the contacts and the sidewalls of the connector is slid axially within the connector to engage the contacts by pressing on their insulative portions, being itself compressed toward the center of the connector by the interactions between cam surfaces on the actuator and on the housing sidewalls.
Another example of this type of connector design is found in U.S. Pat. No. 4,904,197, which was issued to M. K. Cabourne on Feb. 27, 1990, which describes a connector in which multiple rows of contact springs are imbedded in a number of dielectric frame elements adjacently disposed along each side of a central card-receiving slot. A flexible stamped and formed spring extends axially under this slot. These frame elements are either held apart by a central portion of this spring when it is in an upper position, or they are pushed together by the outer portions of this spring when it is lowered. After the daughter card is inserted, a linear cam is pulled axially along the connector, lowering the spring so that pressure is exerted on the contact springs.